Image Processing Device for Re-editable Document Layout Preservation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing document image processing techniques fail to accurately extract reusable data from paper documents, as they do not consider recognition accuracy and often lose the layout and positional relationships, making it difficult for users to effectively reuse character images and convert them into vector data.
Innovation Solution
An image processing device with an analyzing unit for extracting character areas, a character recognition unit for generating character code data, a vectorization unit for creating vector drawing data, and a storage location determination unit to arrange data in an electronic document, allowing for re-editable and accurately positioned vector data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If character recognition is performed without considering recognition accuracy, then processing speed is improved, but data reliability deteriorates
Solution Approach 1:
The patent segments the document processing into distinct stages: character area extraction, recognition processing, and post-processing verification. By dividing the recognition process into manageable segments with intermediate validation steps, the system can maintain high processing speed while ensuring recognition accuracy through targeted verification of critical character areas.
Solution Approach 2:
The patent implements feedback mechanisms where recognition results are validated and verified in post-processing stages. The system uses feedback from recognition confidence levels and validation results to adjust processing parameters, ensuring that high-speed processing does not compromise the reliability of the final recognized data.
2Ease of operation
If document image is divided into separate content areas, then data extraction is improved, but positional relationship information is lost
Solution Approach 1:
The patent employs a nested data structure where character area information is embedded within document structure information, which in turn is embedded within the overall document layout. This nesting preserves hierarchical relationships and positional information at multiple levels, allowing easy extraction of specific content while maintaining the contextual and spatial relationships between different document elements.
Solution Approach 2:
The patent adds dimensional information by storing not only the extracted content data but also the spatial coordinates, layout position, and hierarchical level of each character area. This multi-dimensional data structure enables both easy content extraction and preservation of positional relationships by providing access to information across multiple dimensions (content, position, hierarchy, layout).
3Adaptability or versatility
If only character code data is provided, then data reusability is improved, but layout information is lost
Solution Approach 1:
The patent merges character code data with vector drawing data and layout information into a unified electronic document structure. This combination allows the system to provide machine-readable character codes for text processing while simultaneously preserving the visual layout and positional information through vector graphics, making the data adaptable for both textual and visual reproduction purposes.
Solution Approach 2:
The patent creates a composite data structure that combines multiple data types (character codes, vector graphics, layout information) into a single integrated electronic document format. This composite structure provides the versatility of multiple data formats while maintaining their interrelationships, allowing users to access and utilize different aspects of the data according to their specific needs without losing information.
Data Source
AI summary
Character code data and vector drawing data are both listed and provided in a re-editable manner. Electronic data is generated in which information obtained by vectorizing character areas in an image and information obtained by recognizing characters in the image are stored in respective storage locations. As for the electronic data generated in this manner, because character code data and vector drawing data generated from the input image are both presented by a display and edit program, a user can immediately utilize the both data.